Error Message in obtaining steady state in Dynare

Steady state
#1

I am not able to abtain the sstate for my model. this is the error message i get:

Error using print_info (line 32)
Impossible to find the steady state (the sum of square residuals of the static equations is 14.4437). Either the model
doesn’t have a steady state, there are an infinity of steady states, or the guess values are too far from the solution.

Could someone help

#2

Without the mod-file, further diagnostics are impossible.

Unable to obtain steady state
#3

dsge1.mod (1.8 KB)

this is my model

#4

Check your model. For example,

              y_t = exp(x)*(k_t)^cf1 *(lab_t)^1-cf1;

is missing brackets for the exponent.

#5

Thank you. checked and adjusted the model but still getting the same error message

#6

Always provide the most recent version.

#7

dsge1.mod (1.8 KB)

here is the updated version

#8

From what I can see

                m = m(-1) + e*(t-u)*exp(q);
              i_r = i_r(-1) + (1-u)*exp(q);

implies two unit roots in your model. Thus, there are infinitely many steady states and you cannot let Dynare compute one endogenously. You need to provide one.

#9

Thanks…any suggestions as to how to go about that.

#10

Either get rid of the unit roots or compute the steady state with pencil and paper.

#11

dsge1.mod (1.6 KB) dsge1_steadystate.m (4.0 KB)

Thank you for your time.

I have obtain the steady state analytically and written a steady state file. However, i get an error message of of variables not declare while all variables are indeed declared.

I have also tried providing a steady state model block within the dynare script but it gives me the same error of variables not declared.

I also removed the unit roots but still couldn’t work either

#12

Your steady state file does not contain a valid computation. For example

1-ca1/((p_n)*(c_n)= ca1/(e)*(c_t);

is not a proper Matlab assignment. You cannot use implicit definitions.

If statements within block model
#13

Thank you for your help. I was able to obtain the steady state after rewriting the explicit definitions. I have attached file for you to see if the steady states are correctly computed.

Apart from that I have the following questions:

  1. I want to normalized some variables to one at SS so that i can track the changes. How can i do this on dynare.

  2. I have 3 shocks q,t u. I will only apply shock on q while t and u will take the values of 1 and zero at SS. I know how to apply shock on q but where can i place the t and u such that they will carry the assigned values.

Thanks once again
ds1.mod (2.8 KB)

#14
  1. What do you mean with normalize?
  2. You need to specify a shocks- block.
#15
  1. I want the value of some key variables e.g y, c, p etc to be equal to one in the prior steady state. This will enable me track the changes once i introduce the shock.

Thank you

#16

This requires introducing normalizing constants.

#17

Any suggestion or reference on how to go about introducing normalizing constants.

Also from the mod file above, can i take the first SS as the prior SS and the second as the posterior SS.

#18

Is that even necessary? Can’t you simply express the Variables as a fraction of the initial steady state?

#19

How do you mean specifically?

#20

The first entry of the simulated series will be the old steady state. So you can divide all series by this value to obtain ratios.